1,284 Hits in 7.3 sec

Tutorial on Probabilistic Topic Modeling: Additive Regularization for Stochastic Matrix Factorization [chapter]

Konstantin Vorontsov, Anna Potapenko
2014 Communications in Computer and Information Science  
In this tutorial we introduce a novel non-Bayesian approach, called Additive Regularization of Topic Models.  ...  Probabilistic topic modeling of text collections is a powerful tool for statistical text analysis.  ...  We thank Alexander Frey for his help and valuable discussion, and Vitaly Glushachenkov for his experimental work on model data.  ... 
doi:10.1007/978-3-319-12580-0_3 fatcat:k3buv7ss5rgi7izcgdhseizki4

BigARTM: Open Source Library for Regularized Multimodal Topic Modeling of Large Collections [chapter]

Konstantin Vorontsov, Oleksandr Frei, Murat Apishev, Peter Romov, Marina Dudarenko
2015 Communications in Computer and Information Science  
Tutorial on Probabilistic Topic Modeling: Additive Regularization for Stochastic Matrix Factorization // AIST'2014, Springer CCIS, 2014. Vol. 436. pp. 29-46.  ...  Topic Modeling is an ill-posed inverse problem Topic Modeling is the problem of stochastic matrix factorization: p(w |d) = t∈T φ wt θ td .  ...  Conclusions ARTM (Additive Regularization for Topic Modeling) is a general framework, which makes topic models easy to design, to infer, to explain, and to combine.  ... 
doi:10.1007/978-3-319-26123-2_36 fatcat:ugfml3fa6vhkhjd5suhizd52ye

Renormalization Analysis of Topic Models

Sergei Koltcov, Vera Ignatenko
2020 Entropy  
In this paper, the renormalization procedure is developed for the probabilistic Latent Semantic Analysis (pLSA), and the Latent Dirichlet Allocation model with variational Expectation–Maximization algorithm  ...  In practice, to build a machine learning model of big data, one needs to tune model parameters.  ...  Since for most tasks, it is probabilistic models that are usually applied, here we focus on three most popular probabilistic models: a classical version of pLSA [24] , a classical version of VLDA model  ... 
doi:10.3390/e22050556 pmid:33286328 pmcid:PMC7517079 fatcat:kf3ucghnlzcgtdvco2coipnvoq

RecSim NG: Toward Principled Uncertainty Modeling for Recommender Ecosystems [article]

Martin Mladenov, Chih-Wei Hsu, Vihan Jain, Eugene Ie, Christopher Colby, Nicolas Mayoraz, Hubert Pham, Dustin Tran, Ivan Vendrov, Craig Boutilier
2021 arXiv   pre-print
It offers: a powerful, general probabilistic programming language for agent-behavior specification; tools for probabilistic inference and latent-variable model learning, backed by automatic differentiation  ...  and tracing; and a TensorFlow-based runtime for running simulations on accelerated hardware.  ...  We map these elements to the RecSim NG model library as follows. For interest dynamics, we use a ControlledLinearGaussianStateModel, with an identity transition matrix and as a control matrix.  ... 
arXiv:2103.08057v1 fatcat:wgg3hbvk5nee5fjliwz24rctbm

Mining Ethnic Content Online with Additively Regularized Topic Models

Murat Apishev, Sergei Koltcov, Olessia Koltsova, Sergey Nikolenko, Konstantin Vorontsov
2016 Journal of Computacion y Sistemas  
A recently developed approach to topic modeling, additive regularization of topic models (ARTM), provides fast inference and more control over the topics with a wide variety of possible regularizers than  ...  Topic modeling, additive regularization of topic models, computational social science. and theoretical computer science. Konstantin Vorontsov is the Head  ...  Tutorial on probabilistic topic modeling: Additive regularization for stochastic matrix factorization.  ... 
doi:10.13053/cys-20-3-2473 fatcat:c5exxkkbbrfxtns66poawzh2oi

Advances and challenges of probabilistic model checking

Marta Kwiatkowska, Gethin Norman, David Parker
2010 2010 48th Annual Allerton Conference on Communication, Control, and Computing (Allerton)  
Probabilistic model checking is a powerful technique for formally verifying quantitative properties of systems that exhibit stochastic behaviour.  ...  In this paper, we give a short overview of probabilistic model checking and of PRISM (, currently the leading software tool in this area.  ...  Several different approaches exist to combine both of these factors. One prominent model in this area is probabilistic timed automata (PTAs) [11] , [12] , [13] .  ... 
doi:10.1109/allerton.2010.5707120 fatcat:bauounyehvelpe4dwjrobeugfe

A tutorial on variational Bayes for latent linear stochastic time-series models

Dirk Ostwald, Evgeniya Kirilina, Ludger Starke, Felix Blankenburg
2014 Journal of Mathematical Psychology  
In this tutorial we attempt to provide an introductory overview of the theoretical underpinnings that the variational Bayesian approach to latent stochastic time-series models rests on by discussing its  ...  h i g h l i g h t s • Stochastic time-series modeling. • Linear Gaussian state space models. • Variational Bayes. a b s t r a c t Variational Bayesian methods for the identification of latent stochastic  ...  Linear Gaussian state space models In this section we introduce the central probabilistic model of this tutorial, the linear Gaussian state space model (LGSSM) as an approximation to a latent linear stochastic  ... 
doi:10.1016/ fatcat:ttuhicbanzhuvoyq4uiycvi4gy

Modeling language and cognition with deep unsupervised learning: a tutorial overview

Marco Zorzi, Alberto Testolin, Ivilin P. Stoianov
2013 Frontiers in Psychology  
factorization.  ...  We argue that the focus on deep architectures and generative (rather than discriminative) learning represents a crucial step forward for the connectionist modeling enterprise, because it offers a more  ...  The framework of probabilistic graphical models (Koller and Friedman, 2009 ) provides a general approach to model arbitrarily complex statistical distributions, which can involve a large number of stochastic  ... 
doi:10.3389/fpsyg.2013.00515 pmid:23970869 pmcid:PMC3747356 fatcat:hnszejz7yfeufgsdfbuxktplbe

GemPy 1.0: open-source stochastic geological modeling and inversion

Miguel de la Varga, Alexander Schaaf, Florian Wellmann
2019 Geoscientific Model Development  
In addition, we provide methods to analyze model topology and to compute gravity fields on the basis of the geological models and assigned density values.  ...  The functionality can be separated into the core aspects required to generate 3-D geological models and additional assets for advanced scientific investigations.  ...  Edited by: Lutz Gross Reviewed by: Sally Cripps and one anonymous referee  ... 
doi:10.5194/gmd-12-1-2019 fatcat:xwgacipiljbuldbg47ovnx7ln4

Graphical Models: Foundations of Neural Computation

Michael I. Jordan, Terrence J. Sejnowski
2002 Pattern Analysis and Applications  
The selationship between these components underlies the computational machincq~ associated with graphical models.  ...  1 Bnl/c>sin~r ~ic'tulork, or-the graph may he undircctcci, in \~~I i i c h casc tlic' model is genesally referred to a s a Mur.ko71 rlll~dotri ficllj.  ...  See also Jordan (1999) for several tutorial articles that provide basic background for the chapters presented here.  ... 
doi:10.1007/s100440200036 fatcat:bt75wlwba5hefifledkf62lv4e

A Geometry-Driven Longitudal Topic Model

Yu Wang, Conrad Hougen, Brandon Oselio, Walter Dempsey, Alfred Hero
2021 Harvard data science review  
In addition, the framework permits study of spatial variation in Twitter behavior for learned topics.  ...  A simple and scalable framework for longitudinal analysis of Twitter data is developed that combines latent topic models with computational geometric methods.  ...  The authors thank James Chu for assistance in building the web application for spatiotemporal topic visualization. Contributions.  ... 
doi:10.1162/99608f92.b447c07e doaj:23143f3dd7e449d1be6552936c5d8e55 fatcat:vmvz4j3ax5eohp2hc2pg6kzteq

Dynamic Structural Equation Models for Social Network Topology Inference [article]

Brian Baingana, Gonzalo Mateos, Georgios B. Giannakis
2013 arXiv   pre-print
To infer the network topology, a dynamic structural equation model is adopted to capture the relationship between observed adoption times and the unknown edge weights.  ...  times when blogs mention popular news items, individuals in a community catch an infectious disease, or consumers adopt a trendy electronics product are typically known, and are implicitly dependent on  ...  Several prior approaches postulate probabilistic models and rely on maximum likelihood estimation (MLE) to infer edge weights as pairwise transmission rates between nodes [34] , [27] .  ... 
arXiv:1309.6683v2 fatcat:bzl4j7ceivcnnibq66uw6mtzq4

Hidden Markov Models and their Application for Predicting Failure Events [article]

Paul Hofmann, Zaid Tashman
2020 arXiv   pre-print
The mixtures act as a regularization for typically very sparse problems, and they reduce the computational effort for the learning algorithm since there are fewer distributions to be found.  ...  We show how Markov mixed membership models (MMMM) can be used to predict the degradation of assets. We model the degradation path of individual assets, to predict overall failure rates.  ...  Mixing over common shared distributions acts as a regularization for a typically very sparse problem, thus avoiding overfitting and reducing the computational effort for learning.  ... 
arXiv:2005.09971v1 fatcat:3zvhmwubkvamlmg4dtlovqcmdm

Sparse Solutions to Complex Models [chapter]

Xin Chen, Chung Piaw Teo
2013 Theory Driven by Influential Applications  
To bring to the attention and raise the interest of the operations research community on this topic, we present in this tutorial a wide range of complex models that admit sparse yet effective solutions  ...  Recent years witnessed the proliferation of the notion of sparsity and its applications in operations research models.  ...  He would like to thank Professor Chung Piaw Teo and the department of decision sciences at National University of Singapore for hosting his visit during which this tutorial was prepared.  ... 
doi:10.1287/educ.2013.0116 fatcat:24rbswl2nvb67lnywzobpuyady

Getting Started with Particle Metropolis-Hastings for Inference in Nonlinear Dynamical Models [article]

Johan Dahlin, Thomas B. Schön
2019 arXiv   pre-print
This tutorial provides a gentle introduction to the particle Metropolis-Hastings (PMH) algorithm for parameter inference in nonlinear state-space models together with a software implementation in the statistical  ...  To illustrate the use of PMH, we consider parameter inference in a linear Gaussian state-space model with synthetic data and a nonlinear stochastic volatility model with real-world data.  ...  Acknowledgments This work was supported by the projects: Probabilistic modeling of dynamical systems (Contract number: 621-2013-5524), CADICS, a Linnaeus Center, both funded by the Swedish Research Council  ... 
arXiv:1511.01707v8 fatcat:gapd5ebdgnfzbayf74k5nwkmcy
« Previous Showing results 1 — 15 out of 1,284 results